How Do the Industrial Structure Optimization and Urbanization ... - MDPI

4 downloads 0 Views 2MB Size Report
Jun 5, 2018 - College of Politics and Public Administration, Zhejiang University of ... for Green Low-Carbon Development Research, Zhejiang University of ...
sustainability Article

How Do the Industrial Structure Optimization and Urbanization Development Affect Energy Consumption in Zhejiang Province of China? Huiqin Jiang 1,2,3 , Xiao Zhang 1 , Xinxiao Shao 1 1 2 3

*

ID

and Jianqiang Bao 1,2, *

College of Politics and Public Administration, Zhejiang University of Technology, Hangzhou 310023, China; [email protected] (H.J.); [email protected] (X.Z.); [email protected] (X.S.) Center for Green Low-Carbon Development Research, Zhejiang University of Technology, Hangzhou 310023, China College of Economics and Management, Zhejiang University of Technology, Hangzhou 310023, China Correspondence: [email protected]  

Received: 27 April 2018; Accepted: 4 June 2018; Published: 5 June 2018

Abstract: In response to global climate change, China has voluntarily assumed responsibility and has pledged to reach its peak in carbon emissions by 2030. Industrial structure and urbanization have important impacts on energy consumption. This paper empirically analyzes the dynamic influence of industrial structure and urbanization on energy consumption in the Zhejiang Province of China by constructing a structural vector auto regressive model using impulse response function and variance decomposition. The results show a positive impact of urbanization on energy consumption, which increases and then gradually decreases, and an impact of industrial structure on energy consumption. The results also indicate that it will take a certain period of time for an increase in the proportion of tertiary industry to curb the growth of energy consumption. Urbanization has a greater impact on energy consumption than does industrial structure. Keywords: energy consumption; industrial structure; urbanization; SVAR model

1. Introduction Although China is a developing country, it has assumed responsibility for addressing the problem of global climate change. In July 2015, China officially announced its Intended Nationally Determined Contributions (INDC) and determined its 2030 targets in accordance with its respective national conditions, stages of development, strategy of sustainable development, and international responsibility. One target is that China will reduce its carbon dioxide emissions per unit of GDP by 60–65% of 2005 levels, and non-fossil energy will account for 20% of primary energy consumption [1]. China’s goals have been well received by the international community, and it has contributed to the Paris Agreement. Furthermore, China’s 13th Five-Year Plan for Energy specifically noted that China will implement a dual-control policy of total energy consumption and energy intensity. The plan also clearly indicated that China’s total energy consumption will remain less than 5 billion tons of standard coal, and coal consumption will reach less than 4.1 billion tons by 2020 [2]. China has established a decomposition mechanism for each province; consequently, controlling total energy consumption has been incorporated into the development requirements of the provinces as an important binding target [3]. According to the BP Statistical Review of World Energy 2017 [4], much of the substantial improvement in global carbon emissions from energy consumption can be attributed to changes in the patterns of energy consumption in China. However, this report also revealed that China

Sustainability 2018, 10, 1889; doi:10.3390/su10061889

www.mdpi.com/journal/sustainability

Sustainability 2018, 10, 1889

2 of 12

remained the world’s largest growth market for energy. In 2016, total energy consumption in China amounted to 4.36 billion tons of standard coal, which represents a growth rate of 1.4%, and coal consumption accounted for 62% of total energy consumption. Some studies [5–7] predict that China’s carbon emissions and energy will increase in the coming decades, and almost all of the conclusions claim that China still needs to make even greater changes. Because of the high levels of carbon emissions, coupled with political and technological uncertainties during the process, China should dedicate greater effort to meeting its scheduled emission-reduction targets [8,9]. Providing insight into the impact of energy consumption is essential for analyzing the mechanisms of increased energy use. The relationship between energy consumption and industrial structure changes is an important and challenging issue for policy makers [10,11]. Industrial restructuring, particularly when secondary industry unceasingly declines and tertiary industry rises, may reduce energy intensity and CO2 emission [12–14]. Based on research into the relationship among industrial structure, energy structure, and carbon emission in China, Zhang [15] emphasized that energy-saving emission reduction should occur mainly through improvements in products and energy-saving technologies in secondary industry. Nevertheless, increasing the proportion of tertiary industry is modestly challenging for social resources, energy planning and technological innovation under environmental regulation [16]. Zhu et al. [17] used Tianjin, China as a case study and found that the growing importance of tertiary industry and services (including producer services and consumer services) led to an increase in CO2 emissions because of their growing share of net exports and final demand on the industrial structure. Thus, it is necessary to study the impact of tertiary industry on energy consumption. Many scholars are concerned that, as a corollary of the relationship between urbanization, urban environments and energy consumption are under growing pressure [18–20]. Urbanization is characterized by rapid population transfer from rural to urban areas, and it ameliorates production methods and the lifestyles of residents, but also increases the personal consumption of energy as well as the consumption of energy used for the construction of public infrastructure and for transportation and its infrastructure, which causes a tremendous increase in total energy demand [21–23]. Phetkeo and Shinji [24] used the stochastic impacts by regression on population, affluence, and technology (STIRPAT) model to study the impact of urbanization on energy using the different stages of 99 countries, and their findings suggest that the impact of urbanization on energy use and emissions varies across the stage of development. The negative influence of urbanization on energy consumption has also been studied by some scholars in China [25,26]. However, it has positive impact when the rate of urbanization surpasses 73.8% [27]. With the advancement of urbanization, the concentration of various factors of production has also contributed to the scale of production. Xi et al. [28] compared the ecological environment carrying capacity and the urbanization of both developed and major developing countries and found that cities with more than one million people reported more sustainability of ecological resources; therefore, China should fully consider the economic effect of each metropolis. How do industrial structure optimization and urbanization development affect energy consumption, especially in Zhejiang Province of China, where carbon emissions from energy activities account for approximately 78–80% of total carbon emissions? [29]. The growth rate of urbanization in Zhejiang ranks third out of 34 provinces in China, and it is one of the few provinces in China where tertiary industry has surpassed secondary industry. All of these new challenges have a profound impact on energy consumption and make it much more difficult to realize energy-saving potential. Therefore, we focus on Zhejiang Province and measure the impact of urbanization and industrial structure on energy consumption separately by impulse response function and variance decomposition using a structural vector auto regression (SVAR) model. Quantifying the dynamic response of energy consumption to tertiary industry and urbanization might provide a reference for Zhejiang Province to better understand the characteristics of energy consumption and to enhance the scientific basis of decision-making regarding energy conservation. The policy recommendations made in this paper have

Sustainability 2018, 10, 1889

3 of 12

certain adaptability to some provinces or regions that have been implementing energy conservation and emission reduction for many years, but where the energy saving space is shrinking. The structure of this paper is as follows. In the next section, we explain the energy-saving status of Zhejiang by describing the data. Then the vector auto regression (VAR) model and a SVAR model are described in section three. The fourth section describes the empirical process and results, and3 ofthe final section Sustainability 2018, 10, x FOR PEER REVIEW 12 provides the conclusion. 2. Data

section, we explain the energy-saving status of Zhejiang by describing the data. Then the vector auto regression (VAR) model and a SVAR model are described in section three. The fourth section describes the empirical process and results, and the final section provides the conclusion.

2.1. Energy Consumption Levels 2. Data Energy-saving CO2 emission-reduction efforts in Zhejiang province have achieved remarkable 2.1. Energyand Consumption Levels results over theEnergy-saving past decade. in Figure 1 shows the decoupling and The CO2 polyline emission-reduction efforts in Zhejiang province have elastic achieved coefficient of environmental and energy consumption in The Zhejiang more on this Tapio [30] and remarkable results over the past decade. polyline (for in Figure 1 shows the method, decoupling see elastic ect− (ect − coefficient of environmental and energy in Zhejiang (for more on this method, see 1 ) /ect−consumption 1 Gao [31]). The formula is D f = (ep−ep )/ep , where( ec represents energy consumption, the unit is ten )/ t−1 is = , where ec represents energy Tapio [30] and Gao [31]). Thet−1formula ( )/ thousand tons of standard coal, and ep represents the GDP level (in 1990 constant prices). In accordance consumption, the unit is ten thousand tons of standard coal, and ep represents the GDP level (in 1990 with the classification diagram of Tapio’s index, weoffound Zhejiang province has been constant prices). In accordance with decoupling the classification diagram Tapio’sthat decoupling index, we found that Zhejiang province has been a weak decoupling state since 2006, and from the pressure on consumption in a weak decoupling state since 2006, and in the pressure on the environment energy the environment from energy consumption has been weakened. However, total energy consumption has been weakened. However, total energy consumption continues to steadily increase. In 2016, energy continues to steadily increase. In 2016, energy consumption in Zhejiang reached 2.03 million tons of consumption inand Zhejiang reached tons of coal,higher and than the the growth 3.4%, which was coal, the growth rate was2.03 3.4%,million which was significantly overallrate level was of 1.4% in significantlyChina. higher than the overall level of 1.4% in China. 25,000

1.2 1

20,000

0.8 15,000 0.6 10,000 0.4 5,000

0.2 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

0

Energy Consumption Decoupling index 1. Elastic coefficient and consumption of decoupling Zhejiang Province. The Figure 1. Figure Elastic coefficient andenergy energy consumption ofof decoupling ofNote: Zhejiang Province. original data were obtained from the Zhejiang Statistical Yearbook [32] and the China Energy Statistics Note: The original data were obtained from the Zhejiang Statistical Yearbook [32] and the China Energy Yearbook [33]. Statistics Yearbook [33].

2.2. Industrial Structure The proportion 2.2. Industrial Structure

of secondary industry has successively declined since 2009 in Zhejiang Province, while the tertiary industry simultaneously experienced accelerated growth and surpassed secondary industry in 2014, forming a “321” structuredeclined pattern (see Figure 2). The of The proportion of secondary industry hasindustrial successively since 2009 inlevel Zhejiang Province, industrial structure rationalization in Zhejiang Province is currently quite reasonable [34], although while the tertiary industry simultaneously experienced accelerated growth and surpassed secondary it may experience greater resistance to industrial restructuring [35]. The growth of tertiary industry industry in brought 2014, forming a “321” industrial structure Figure 2).this The level about by changes in energy consumption is also pattern worthy of (see further study. In paper, we of industrial

structure rationalization in Zhejiang Province is currently quite reasonable [34], although it may experience greater resistance to industrial restructuring [35]. The growth of tertiary industry brought about by changes in energy consumption is also worthy of further study. In this paper, we use data on the proportion of tertiary industries to GDP to represent the optimization of industrial structure, which is expressed as is.

Sustainability 2018, 10, x FOR PEER REVIEW

4 of 12

use data on the proportion of tertiary industries to GDP to represent the optimization of industrial4 of 12 structure, which is expressed as is.

Sustainability 2018, 10, 1889

Sustainability 2018, 10, x FOR PEER REVIEW

4 of 12

use data on the proportion of tertiary industries to GDP to represent the optimization of industrial structure, which is expressed as is.

Figure 2. Proportional contributiontotothe theGDP GDP of of Zhejiang Zhejiang Province Note: The The original Figure 2. Proportional contribution Province(1990–2016). (1990–2016). Note: original data were obtained from ZhejiangStatistical StatisticalYearbook. Yearbook. data were obtained from thethe Zhejiang

2.3. Level of Urbanization

Figure 2. Proportional contribution to the GDP of Zhejiang Province (1990–2016). Note: The original 2.3. Level of Urbanization obtained from there the Zhejiang Yearbook. large population changes in Zhejiang, and In thedata lastwere several years, haveStatistical been relatively

In several years, havetobeen Zhejiang, and the thethe ratelast of urbanization hasthere increased 67%relatively (see Figurelarge 3). Inpopulation the last two changes years, for in instance, a large 2.3. Level of Urbanization rate of urbanization has increased to 67% (see Figure 3). In the last two years, for instance, a large number of people poured into Hangzhou, the provincial capital city, and its population increased by In the last several years, there have been relatively large population changes in Zhejiang, and number of people poured into Hangzhou, the provincial capital city, and its population housing, increased by 296,000. The acceleration of urbanization places greater demands on urban transportation, the rate of urbanization has increased to 67% (see Figure 3). In the last two years, for instance, a large 296,000. acceleration ofhousehold urbanization places greater demands on urban and The materials, as well as energy consumption, and these issues placetransportation, a great strain onhousing, the number of people poured into Hangzhou, the provincial capital city, and its population increased by carrying capacity of urban resources and the environment. To enhance the accuracy of the data, and materials, as well as household energy consumption, and these issues place a great strain 296,000. The acceleration of urbanization places greater demands on urban transportation, housing, the on urbanization data for all years up to 2006 areconsumption, the thatand were modified by Zhou Tian The data, andcapacity materials, as well as household energy these issues place a great strain on [36]. the the carrying of urban resources and thedata environment. To enhance theand accuracy of the data for all other years were obtained from the Zhejiang Statistical Yearbook and the China carrying capacity of urban resources and the environment. To enhance the accuracy of the data, the the urbanization data for all years up to 2006 are the data that were modified by Zhou and Tian [36]. Statistical Yearbook. level of up urbanization is represented u and is as a[36]. percentage. urbanization dataThe for all years to 2006 are the data that wereby modified byexpressed Zhou and Tian The The data for all other years were obtained from the Zhejiang Statistical Yearbook and the China data for all other years were obtained from the Zhejiang Statistical Yearbook and the China Statistical Statistical Yearbook. The level of urbanization u and is expressed as a percentage. Yearbook. The level of urbanizationisisrepresented represented byby u and is expressed as a percentage. 0.80

0.600.70 Urbanization Level

Urbanization Level

0.700.80 0.500.60 0.400.50 0.300.40 0.30

0.20

0.20

0.10

0.10

0.00

0.00

Year Year Figure 3. Urbanization rate of Zhejiang Province (1990–2016).

Figure Urbanizationrate rate of of Zhejiang (1990–2016). Figure 3. 3. Urbanization ZhejiangProvince Province (1990–2016).

3. Methodology 3.1. SVAR Model Although industrial structure and urbanization affect energy consumption differently, they both significantly affect energy consumption. Based on the above review of the opinions of scholars (Chen and Lei, 2017; Xiao et al., 2017; Phetkeo and Shinji, 2010; Liu et al., 2017) and considering the level of energy consumption in Zhejiang Province, we found that an effective energy consumption

Sustainability 2018, 10, 1889

5 of 12

model should include variables for industrial structure and urbanization. A more useful method for determining the endogenous and exogenous variables in a modeling process was to build an unconstrained multivariate time series model instead of a static regression model. Zellner [37] noted that any simultaneous equations modeled with one or more endogenous lag variables will eventually lead to a model that treats each exogenous variable in the system as a function of the lagged value of the endogenous variable model. Sims [38] proposed the vector autoregressive model in 1980, which has subsequently been commonly used in econometrics. This type of model can support a macroeconomic structural analysis. Therefore, we considered a general VAR model: Yt = A1 Yt−1 + . . . + A p Yt− p + ε t ,

(1)

where p is the lag length, and ε t is the stochastic disturbance term or interest. However, the VAR model has some limitations in describing the contemporaneous relationships among the variables; it hides the contemporaneous correlation to ε t , which makes it impractical for the impulse response function to explain the economic meaning. The monograph of Amisano and Giannini [39] was, to a certain extent, of monumental significance to the SVAR model because these two econometricians in their monograph summarized the establishment, identification, estimation, and application of SVAR models. The SVAR model not only depicts the influence of each variable on each lag item but also captures the instantaneous structural relationship among variables in the model system. In recent studies, the authors further validated and expanded the SVAR, and the identification conditions were further explained [40,41]. The expression of SVAR is: B0 Yt = Φ1 Yt−1 + . . . + Φ p Yt− p + µt ,  where B0 

  =   

1 −b21 .. . −bk1

−b12 1 .. . −bk2

· · · −b1k . . . −b2k .. .. . . ··· 1

(2)

   , which can reflect the contemporaneous correlation,  



µ1,t    µ2,t   µt =  .  . Considering the lag between the impact of the variables, we introduced a hysteresis  ..  µk,t operator L and express it as: B( L)Yt = µt (3) B(L) = B0 − Φ1 L − Φ2 L2 − . . . − ΦP L p , which was also expressed as a moving average form Yt = D ( L)µt

(4)

Since D(L) = B( L)−1 = D0 + D1 L + D2 L2 + . . . , we can obtain D0 = B0 −1 The final expression of the SVAR model was: C ( L)ε t = D ( L)µt ,

(5)

where C(L) and D(L) are lag variables for the VAR model and the SVAR model, respectively. If we want to obtain the estimated number of the VAR model that is not less than the estimated amount of the SVAR model, it is necessary to impose certain constraints, which involves the use of the recognition process of the SVAR model. Constraints should be based on reality and theory [42].

Sustainability 2018, 10, 1889

6 of 12

3.2. ADF Test, Lag Order Determination and Model Stationarity Test Ensuring the applicability of the data is critical in econometric analysis. Variations in the economic structure caused by irregular economic growth, the discontinuity of energy policy and technical instability could cause most energy-related variables to fluctuate. To eliminate the heteroscedasticity of time series data and to facilitate the economic interpretation of the results, we used the logarithmic function for certain data: lnec, lnu, and lnis. The traditional VAR model required that all variables must beSustainability stationary. Inxaddition, differential data processing was the most commonly used method, 2018, 10, FOR PEER REVIEW 6 of 12 but some sequential information may be lost. Gao [43] stressed that the VAR model can also be heteroscedasticity of time series data and to facilitate the economic interpretation of the results, we established directly as long as the variables are cointegrated, even if the time series is nonstationary. used the logarithmic function for certain data: lnec, lnu, and lnis. The traditional VAR model required Therefore, a cointegration on thedifferential sequencedata andprocessing used the was trace that we all performed variables must be stationary. Intest addition, thestatistic most to inspect and judge relationships of the variables. The results showed that the hypothesis of commonly used method, but some sequential information may be lost. Gao [43] stressed thatthere the existing at VAR model can also be established directly as long as the variables are cointegrated, even if the time most two cointegrated relationship was very significant at 0.05 level (Prob. = 0.0130). series is nonstationary. Therefore, we performed a cointegration test on the sequence and used the Next, the lag order must be determined. Table 1 shows the results for the VAR lag order selection trace statistic to inspect and judge relationships of the variables. The results showed that the in termshypothesis of the criterion diverse information. By using the lag of LR, FPE, of there for existing at most two cointegrated relationship wascriteria very significant at 0.05AIC, level SC, and HQ, we set the lag= order (Prob. 0.0130).of the VAR model at 5 because three criteria selected the 5th order. Next, the lag order must be determined. Table 1 shows the results for the VAR lag order selection in terms of the criterion fororder diverse information. By using lag Model. criteria of LR, FPE, AIC, Table 1. Lag selection criteria of thethe VAR SC, and HQ, we set the lag order of the VAR model at 5 because three criteria selected the 5th order.

Lag

LogL

LR

PEP

AIC

Table 1. Lag order selection criteria of the VAR Model.

SC

HQ

0 82.22521 NA −7.202292 −7.053514 −7.167244 1.50 × 10−7 PEP AIC SC HQ −10 1 Lag 170.1423LogL 143.8643LR −14.37657 −13.78146 −14.23638 1.16 × 10 82.2252122.70958 NA 1.50 × 11 10−7 −−7.202292 −7.053514 2 0 186.7960 * 15.07236 −14.03091 −7.167244 −14.82703 6.12 × 10− 170.14238.595984 143.8643 7.84 1.16 10−10 −−14.37657 −13.78146 3 1 194.6756 14.97051 −13.48273 −14.23638 −14.62003 × 10×−11 186.796013.73203 22.70958 * 5.29 6.12 10−11 −−15.07236 −14.03091 4 2 211.4592 15.67811 −13.74399 −14.82703 −15.22249 × 10×−11 194.675612.56129 8.5959842.82 ×7.84 −14.97051* −13.48273 5 3 234.4882 −14.57302 * −14.62003 −16.39271 * 10−×1110*−11 −16.95348 211.4592 13.73203 Ratio5.29 10−11 (each −15.67811 Note: LR =4sequential modified Likelihood test×statistic test at 5%−13.74399 level); FPE =−15.22249 Final prediction error −11 * 5 234.4882 12.56129 2.82 × 10 −16.95348HQ * =−14.57302 * −16.39271 * AIC = Akaike information criterion; SC = Schwarz information criterion; Hannan-Quinn information criterion. Note: = sequential modified Likelihood Ratio test statistic (each test at 5% level); FPE = Final * indicates lag LR order selected by the criterion. prediction error AIC = Akaike information criterion; SC = Schwarz information criterion; HQ = Hannan-Quinn indicates lag order selected by the criterion.the results of the test of model A stationary test ofinformation the VARcriterion. model*was conducted. Figure 4 shows stationarity A bystationary AR root,test and VAR inverse of the AR characteristic of clearly, the VAR the model wasmodel’s conducted. Figureroots 4 shows the results of the test of polynomial model stationarity by AR root, and clearly, the VAR model’s inverse roots of the AR characteristic were all within the unit circle, which satisfied stationarity. Therefore, the VAR model can be further polynomial were all within the unit circle, which satisfied stationarity. Therefore, the VAR model developed into the SVAR model.

can be further developed into the SVAR model. 1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Figure 4.4.Test of of model stationarity. Figure Test model stationarity.

3.3. Identification of Constraints

3.3. Identification of Constraints

We constructed an SVAR model with three endogenous variables (i.e., k = 3), imposing

constraints by the recursive Cholesky. In this paper, we chose the AB-matrix Weshort-term constructed an SVAR model form withofthree endogenous variables (i.e., k = 3), imposing because the characteristics AB- can clearly establish the contemporaneous structure short-term constraints by theofrecursive form of Cholesky. In this paper, werelationship chose the AB-matrix of each endogenous variable in the system and can also intuitively analyze the influences of because the characteristics of AB- can clearly establish the contemporaneous structure relationship

Sustainability 2018, 10, 1889

7 of 12

of each endogenous variable in the system and can also intuitively analyze the influences of orthonormal random shocks on the system after impact. Additionally, the AB-matrix requires at least 2k2 − k(k + 1)/2 = 12 constraints to make the model recognizable. In Formula (2), the diagonal element of matrix B0 was 1 (i.e., matrix C in Formula (5)). According to [41], a uniquely identified model 2018,set 10, xof FOR PEER REVIEW 12 requiresSustainability a sufficient restrictions on matrix D, so we further assumed that there was no7 ofcorrelation between the structural perturbation terms in the SVAR model; therefore, matrix D was a diagonal matrix. orthonormal random shocks on the system after impact. Additionally, the AB-matrix requires at Theleast constraints the matrix effectively constrained the correlations 2 (k +imposed 1)/2 = 12 on constraints to make the model recognizable. Incontemporaneous Formula (2), the diagonal among element the variables. that constraints should be determined the economic of matrix Watson was 1suggested (i.e., matrix C in Formula (5)). According to [41], a uniquelyby identified modelon requires sufficient setisofbased. restrictions on matrixenergy-ecology D, so we further assumed that there was no implications whicha the model Regrettably, is a complex system, and there the structural perturbation terms in the SVAR therefore, matrix Din was a paper. is not acorrelation perfectlybetween applicable economic theory that explains the model; constraints proposed this diagonal matrix. Therefore, we referred to previous studies and the basic influence mechanisms of the variables The constraints imposed on the matrix effectively constrained the contemporaneous when determining the constraints. Apparently, changes in urbanization are mostly caused by correlations among the variables. Watson suggested that constraints should be determined by the demographic changes [21,22], whilethe transformations industrial structure are caused by economic implications on which model is based. in Regrettably, energy-ecology is aoften complex policies,system, investments, technology [11,15,16]. a time lagexplains in the the impact of industrial and thereand is not a perfectly applicable Moreover, economic theory that constraints proposed in this paper. Therefore, we out referred to previous and appears the basictoinfluence structure on urbanization cannot be ruled because tertiarystudies industry be expanding of the variables when determining constraints. Apparently, changes in urbanization rapidly mechanisms and disorderly, and it cannot match thetheurbanization plan in China. Therefore, based on are mostly caused by demographic changes [21,22], while transformations in industrial structure are the influence transmission mechanism of urbanization-industrial institutions-energy consumption often caused by policies, investments, and technology [11,15,16]. Moreover, a time lag in the impact (see Figure 5), the constraints were as follows: (1) Urbanization does not respond to changes in of industrial structure on urbanization cannot be ruled out because tertiary industry appears to be contemporaneous energy consumption, C21 =match 0; (2) the theurbanization industrial structure doesTherefore, not respond to expanding rapidly and disorderly, andand it cannot plan in China. changesbased in contemporaneous consumption, and = 0; and (3) urbanization does not respond on the influence energy transmission mechanism of Curbanization-industrial institutions-energy 31 consumption (see Figure 5), the constraints were as follows: (1) Urbanization does not respond to to changes in the contemporaneous industrial structure, and C23 = 0. The results of the parameter changes in contemporaneous energy consumption, and = 0; (2) the industrial structure does not estimation of the SVAR model were expressed as a matrix, as follows: respond to changes in contemporaneous energy consumption, and = 0; and (3) urbanization  to changes in the contemporaneous   industrial structure, and  = 0.The results of does not respond 1 −0.72 −0.32 εˆ 1t 0.01 0 0 µˆ 1t the parameter were as a matrix, as  follows:  estimation of the SVARmodel   expressed   

 0 0

1 1 −0.22 0 0

0 0.01 0  µˆ 2t   εˆ 2t  =  0 ̂ 0.01 0 0 0.72 0.32 ̂ ˆ 11 ε 0 0 0.02 µˆ 3t 3t ̂ ̂ 0 0.01 0 0 0.22

1

̂

0

0

0.02

̂

5. Influence transmission mechanism of urbanization-industrial institutions-energy Figure Figure 5. Influence transmission mechanism of urbanization-industrial institutions-energy consumption. consumption.

Sustainability FOR PEER REVIEW Sustainability 2018, 2018, 10, 10, x1889 Sustainability 2018, 10, x FOR PEER REVIEW

88 of of 12 12 8 of 12

4. Findings 4. Findings 4. Findings The orthogonalized impulse response function could be obtained by using the SVAR model. The orthogonalized impulse response function could be obtained by using the SVAR model. The orthogonalized impulse response function be on obtained by variables. using the In SVAR model. Therefore, we separately analyzed the impact of each could variable the other this paper, Therefore, we separately analyzed the impact of each variable on the other variables. In this paper, Therefore, analyzedfunction the impact of each on the other variables. In this paper, we selectedwe theseparately impulse response of lag 20 to variable analyze the dynamic impact of urbanization we selected the impulse response function of lag 20 to analyze the dynamic impact of urbanization we selected thestructure impulse on response of lag(see 20 toFigures analyze the dynamic impact of urbanization and industrial energyfunction consumption 6 and 7). The variance decomposition and industrial structure on energy consumption (see Figures 6 and 7). The variance decomposition and used industrial structure on (see Figures 6 and The variance decomposition was to characterize theenergy relativeconsumption importance of urbanization and7). industrial structure for energy was used to characterize the relative importance of urbanization and industrial structure for energy was used to characterize relative consumption (see Figuresthe 8 and 9). importance of urbanization and industrial structure for energy consumption (see Figures 8 and 9). consumption Figures 8 and 9). of energy consumption to an urbanization shock. The positive Figure 6 (see depicts the response Figure 6 depicts the response of energy consumption to an urbanization shock. The positive Figure 6 depicts the of energy consumption to an urbanization shock.and The positive impact of urbanization onresponse energy consumption in the early period increased rapidly peaked in impact of urbanization on energy consumption in the early period increased rapidly and peaked in impact of urbanization on energy consumption in the early period increased rapidly and peaked in the the eighth period, which was in line with the actual situation. The increase in the population caused the eighth period, which was in line with the actual situation. The increase in the population caused eighth period, which in line with the actual situation. in the population caused by by urbanization will was clearly cause a considerable increaseThe in increase energy demand in the early stage. by urbanization will clearly cause a considerable increase in energy demand in the early stage. urbanization will clearlyof cause a considerable in energy in allocation the early stage. Increasing Increasing the quality urbanization and increase increasing urbandemand resource efficiency can Increasing the quality of urbanization and increasing urban resource allocation efficiency can the quality urbanization and increasing urbanconsumption resource allocation efficiency impact weaken theofimpact of urbanization on energy and may cause can the weaken impacts the to slowly weaken the impact of urbanization on energy consumption and may cause the impacts to slowly of urbanization on energy may cause the impacts to slowly weaken in period, the future [44]. weaken in the future [44].consumption However, theand sustained impact appeared a relatively long which weaken in the future [44]. However, the sustained impact appeared a relatively long period, which However, the sustained appeared a of relatively long period, which poses a challenge the target poses a challenge to the impact target constraints energy consumption for a provincial energyto planning. poses a challenge to the target constraints of energy consumption for a provincial energy planning. constraints energy consumption for aofprovincial planning. Figure of 7 indicates that the impact industrialenergy structure on energy consumption was positive Figure 7 indicates that the impact of industrial structure on energy consumption was positive impact of industrial energyand consumption then then Figure became7 indicates negative.that Thethe impact increased fromstructure the firstonperiod decreasedwas afterpositive the fourth then became negative. The impact increased from the first period and decreased after the fourth became negative. The impact increased from the first period and decreased after the fourth period, period, becoming negative in the eighth period and continuing in this pattern. The results could be period, becoming negative in the eighth period and continuing in this pattern. The results could be becoming negative in the period in and continuing in this pattern. results could be explained explained by the fact thateighth an increase the tertiary industry did notThe immediately reduce energy explained by the fact that an increase in the tertiary industry did not immediately reduce energy by the fact thatOne an increase in the tertiary did not immediately reduce energy consumption. consumption. of the main reasons is industry the heterogeneity of tertiary industry, the development of consumption. One of the main reasons is the heterogeneity of tertiary industry, the development of One of thesuch mainas reasons is the heterogeneity of postal tertiaryservices industry, the development of industries industries transportation, storage and caused the continued growth of such coal industries such as transportation, storage and postal services caused the continued growth of coal as transportation, storage postal services thenegative continued growth of coal oil and other and oil and other types of and high-carbon energycaused use. The impact of the laterand period indicates and oil and other types of high-carbon energy use. The negative impact of the later period indicates types high-carbon use. The negativestructure impact ofwill the eventually later periodinhibit indicates an adjustment that anofadjustment in energy the long-term industrial the that increase in energy that an adjustment in the long-term industrial structure will eventually inhibit the increase in energy in the long-term industrial structure will eventually inhibit the increase in energy consumption. consumption. consumption. .06 .06 .05 .05 .04 .04 .03 .03 .02 .02 .01 .01 .00 .00 -.01 -.01

2 2

4 4

6 6

8 8

10 10

12 12

14 14

16 16

18 18

20 20

Figure 6. Response of energy consumption to urbanization. Figure Figure 6. 6. Response Response of of energy energy consumption consumption to to urbanization. urbanization. .06 .06 .04 .04 .02 .02 .00 .00 -.02 -.02 -.04 -.04

2 2

4 4

6 6

8 8

10 10

12 12

14 14

16 16

18 18

20 20

Figure 7. 7. Response Responseof ofenergy energy consumption consumption to to industrial industrial structure. structure. Figure Figure 7. Response of energy consumption to industrial structure.

Figures 8 and 9 show that the contributions of the shocks of urbanization and industrial Figures 8 and 9 show that the contributions of the shocks of urbanization and industrial structure to the energy consumption. After the first lag period, the main contribution of urbanization structure to the energy consumption. After the first lag period, the main contribution of urbanization

Sustainability 2018, 10, 1889

9 of 12

Figures 8 and show that the contributions Sustainability 2018, 10, PEER REVIEW Sustainability 2018, 10,xx9FOR FOR PEER REVIEW

of the shocks of urbanization and industrial structure 99of of12 12 to the energy consumption. After the first lag period, the main contribution of urbanization shock shock quickly, and to since tenth period. The shock contributors increased quickly, and tended to stabilize since the the tenth laglag period. The shockincreased increased quickly, andtended tended tostabilize stabilize since the tenth lag period. Theshock shockcontributors contributorsof of industrial structure were slightly lower than for urbanization, the impacts of industrial structure structure were slightly lower than for urbanization, the impacts of industrial structure shock industrial structure were slightly lower than for urbanization, the impacts of industrial structure shock slow tenth period. Zhejiang pay showed slow modifications after theafter tenththe period. Therefore, Zhejiang should payshould more attention to shock showed showed slow modifications modifications after the tenth period. Therefore, Therefore, Zhejiang should pay more more attention to the impact when making plans regarding energy the impact urbanization when making plans regarding energy consumption. attention toof the impactof ofurbanization urbanization when making plans regarding energyconsumption. consumption. 100 100 80 80 60 60 40 40 20 20 00

22

44

66

88

10 10

12 12

14 14

16 16

18 18

20 20

Figure Figure8. 8.Percent Percentof ofenergy energyconsumption consumptionvariance variancedue dueto tourbanization. urbanization. Figure 8. Percent of energy consumption variance due to urbanization. 100 100 80 80 60 60 40 40 20 20 00

22

44

66

88

10 10

12 12

14 14

16 16

18 18

20 20

Figure 9. Percent of energy consumption variance due to industrial structure. Figure Figure9.9.Percent Percentof ofenergy energyconsumption consumptionvariance variancedue dueto toindustrial industrialstructure. structure.

5. Conclusions Conclusions 5. 5. Conclusions Empirical results results showed that that urbanization has has caused aa significant significant increase in in energy Empirical Empirical results showed showed that urbanization urbanization has caused caused a significant increase increase in energy energy consumption, which poses a certain challenge to the layout of urban ecological space. In the process of consumption, which poses a certain challenge to the layout of urban ecological space. In the process consumption, which poses a certain challenge to the layout of urban ecological space. In the process urbanization, in addition to the use of radiation in the metropolis, the energy useenergy mode of microbodies of of urbanization, urbanization, in in addition addition to to the the use use of of radiation radiation in in the the metropolis, metropolis, the the energy use use mode mode of of should also be a focus. The energy-saving potential of energy consuming subjects, such as the microbodies microbodiesshould shouldalso alsobe beaafocus. focus.The Theenergy-saving energy-savingpotential potentialof ofenergy energyconsuming consumingsubjects, subjects,such such public sectors and families families remains to be tapped. However, However, it may be difficult be for the government as as the the public public sectors sectors and and families remains remains to to be be tapped. tapped. However, itit may may be difficult difficult for for the the to exert its influence on influence families. Family consumption patterns thatpatterns are related toare urbanization government to exert its on families. Family consumption that related government to exert its influence on families. Family consumption patterns that are related to to need guidance andguidance improvement, such as actively building low-carbon transportation, developing urbanization need and improvement, such as actively building low-carbon transportation, urbanization need guidance and improvement, such as actively building low-carbon transportation, low-carbon communities, advocating for low-carbon travel, green travel, living, and soliving, on. The public sector, developing developinglow-carbon low-carboncommunities, communities,advocating advocatingfor forlow-carbon low-carbon travel,green green living,and andso soon. on.The The including government and schools, can schools, help raise awareness in the energy market and in energy public publicsector, sector,including includinggovernment governmentand and schools,can canhelp helpraise raiseawareness awarenessin inthe theenergy energymarket marketand and enterprises regarding changes in energy demand caused by urbanization by cooperating with industry in in energy energy enterprises enterprises regarding regarding changes changes in in energy energy demand demand caused caused by by urbanization urbanization by by cooperating cooperating and helping them construct low-carbon production facilities in a timely mannera to improve energy with with industry industry and and helping helping them them construct construct low-carbon low-carbon production production facilities facilities in in a timely timely manner manner to to efficiency and achieve low-carbon urban development. improve improveenergy energyefficiency efficiencyand andachieve achievelow-carbon low-carbonurban urbandevelopment. development. The response response of energy energy consumption to to the optimization optimization of industrial industrial structure was was relatively The The response of of energy consumption consumption to the the optimization of of industrial structure structure was relatively relatively large in in the early early periods,and and theincrease increase intertiary tertiary industry will eventually curb the growth large large in the the early periods, periods, and the the increase in in tertiary industry industry will will eventually eventually curb curb the the growth growth of of of energy consumption. Zhejiang Province is at the forefront of China’s provinces by optimizing energy consumption. Zhejiang Province is at the forefront of China’s provinces by optimizing and energy consumption. Zhejiang Province is at the forefront of China’s provinces by optimizing and and upgrading its industrial structure. However, an increase in the proportion of the tertiary upgrading upgrading its its industrial industrial structure. structure. However, However, an an increase increase in in the the proportion proportion of of the the tertiary tertiary industry industry industry does not necessarily meanthe that the industrial structure has been optimizedororupgraded. upgraded. For does does not not necessarily necessarily mean mean that that the industrial industrial structure structure has has been been optimized optimized or upgraded. For For example, the “cost-sickness” that may result from the development of the service sector will affect example, the “cost-sickness” that may result from the development of the service sector will affect the theoverall overallimprovement improvementin intotal totalfactor factorproductivity. productivity. Therefore, Therefore,ititisisespecially especiallyimportant importantto todevelop develop and and improve improve the the efficiency efficiency of of tertiary tertiary industry. industry. The The trend trend of of industrial industrial structure structure evolution evolution under under environmental constraints is bound to be a transformation that includes low-carbon technology. environmental constraints is bound to be a transformation that includes low-carbon technology. At At

Sustainability 2018, 10, 1889

10 of 12

example, the “cost-sickness” that may result from the development of the service sector will affect the overall improvement in total factor productivity. Therefore, it is especially important to develop and improve the efficiency of tertiary industry. The trend of industrial structure evolution under environmental constraints is bound to be a transformation that includes low-carbon technology. At present, the technology potential of Zhejiang’s tertiary industry has not yet been realized. Therefore, sustained and effective policy support can prompt the rapid growth of low-carbon energy technologies from a niche market and achieve environmental benefits as soon as possible. However, as traditional industries transform and upgrade, the fairness and adaptability of their policies should also be considered in the process. Author Contributions: H.J. and J.B. conceived and designed the study. X.Z. and X.S. processed the data and performed the experiments. H.J. and X.Z. reviewed and edited the manuscript. All authors read and approved the manuscript. Funding: This research is supported by the Natural Science Foundation of Zhejiang Province (No. LQ15G030008) and the National Social Science Foundation of China (No. 15BGL169). Conflicts of Interest: The authors declare no conflict of interest.

References 1.

2.

3. 4.

5. 6. 7. 8. 9. 10. 11.

12. 13. 14.

Intended Nationally Determined Contributions (INDC). National Development and Reform Commission. Intended Nationally Determined Contributions. The State Council Information Office of the People’s Republic of China. 2017. Available online: http://www.scio.gov.cn/xwfbh/xwbfbh/wqfbh/2015/20151119/ xgbd33811/Document/1455864/1455864.htm (accessed on 12 April 2017). Energy Development “13th Five-Year Plan”. National Development and Reform Commission & National Energy Administration. 2016. Available online: http://www.ndrc.gov.cn/zcfb/zcfbtz/201701/t20170117_ 835278.html (accessed on 26 December 2016). Liu, Y.; Xiao, H.; Lv, Y.; Zhang, N. The effect of new-type urbanization on energy consumption in China: A spatial econometric analysis. J. Clean. Prod. 2017, 163, S299–S305. [CrossRef] BP. BP Statistical Review of World Energy. 2017. Available online: https://www.bp.com/content/dam/ bp/en/corporate/pdf/energy-economics/statistical-review-2017/bp-statistical-review-of-world-energy2017-full-report.pdf (accessed on 10 December 2017). Yuan, X.C.; Wei, Y.M.; Mi, Z.; Sun, X.; Zhao, W.; Wang, B. Forecasting China’s regional energy demand by 2030: A Bayesian approach. Resour. Conserv. Recycl. 2017, 127, 85–95. [CrossRef] Xiao, B.; Niu, D.; Wu, H. Exploring the impact of determining factors behind CO2 emissions in China: A CGE appraisal. Sci. Total Environ. 2017, 581, 559–572. [CrossRef] [PubMed] Wang, J.; Li, L. Sustainable energy development scenario forecasting and energy saving policy analysis of China. Renew. Sustain. Energy Rev. 2016, 58, 718–724. [CrossRef] Andrews Speed, P. China’s Ongoing energy energy efficiency drive: Progress and prospects. Energy Policy 2009, 37, 1331–1344. [CrossRef] Kahrl, F.; Roland-Holst, D. Growth and structural change in China’s energy economy. Energy 2009, 34, 894–903. [CrossRef] Juan Antonio, D.; Alcántara, V.; Padilla, E. International inequality in energy intensity levels and the role of production composition and energy efficiency an analysis of OECD countries. Ecol. Econ. 2010, 69, 2468–2474. Li, Z.; Sun, L.; Geng, Y.; Dong, H.; Ren, J.; Liu, Z.; Tian, X.; Yabar, H.; Higano, Y. Examining industrial structure changes and corresponding carbon emission reduction effect by combining input-output analysis and social network analysis: A comparison study of China and Japan. J. Clean. Prod. 2017, 162, 61–70. [CrossRef] Li, K.; Lin, B.Q. The nonlinear impacts of industrial structure on China’s energy intensity. Energy 2014, 69, 258–265. [CrossRef] Chen, W.H.; Lei, Y.L. Analysis of the impact path on factors of China’s energy-related CO2 emissions: A path analysis with latent variables. Environ. Sci. Pollut. Res. 2017, 24, 5757–5772. [CrossRef] [PubMed] Fisher, K.; Jefferson, G.; Liu, H.; Tao, Q. What is driving China’s decline in intensity? Resour. Energy Econ. 2004, 26, 77–97. [CrossRef]

Sustainability 2018, 10, 1889

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

11 of 12

Zhang, L. Relations among the industry structure, energy structure and carbon emissions. J. Arid Land Resour. Environ. 2011, 5, 1–7. Jorgenson, D.W.; Timmer, M.P. Structural Change in Advanced Nations: A New Set of Stylised Fact. Scand. J. Econ. 2011, 113, 1–29. [CrossRef] Zhu, Z.; Liu, Y.; Tian, X.; Wang, Y.; Zhang, Y. CO2 emissions from the industrialization and urbanization processes in the manufacturing center Tianjin in China. J. Clean. Prod. 2017, 168, 867–875. [CrossRef] Parikh, J.; Shukla, V. Urbanization, energy use and greenhouse effects in economic development. Results from a cross-national study of development countries. Fuel Energy Abstr. 1995, 5, 87–103. Martínez-Zarzoso, I.; Maruotti, A. The Impact of Urbanization on CO2 Emissions: Evidence from Developing Countries. Ecol. Econ. 2011, 70, 1344–1353. [CrossRef] Al-Mulali, U.; Che, N.; Fereidouni, H. Exploring the bi-directional long run relationship between urbanization, energy consumption, and carbon dioxide emission. Energy 2012, 46, 156–167. [CrossRef] Liddle, B. Impact of population, age structure, and urbanization on carbon emissions/energy consumption: Evidence from macrolevel, cross-country analyses. Popul. Environ. 2014, 35, 286–304. [CrossRef] Otsuka, A. Population Agglomeration and Residential Energy Consumption: Evidence from Japan. Sustainability 2018, 10, 469. [CrossRef] Reinhard, M.; Yasin, S. Impacts of urbanization on urban structures and energy demand: What can we learn for urban energy planning and urbanization management? Sustain. Cities Soc. 2011, 1, 45–53. Phetkeo, P.; Shinji, K. Does urbanization lead to less energy use and lower CO2 emissions? A cross-country analysis. Ecol. Econ. 2010, 70, 434–444. Fan, J.L.; Zhang, Y.J.; Wang, B. The impact of urbanization on residential energy consumption in China, An aggregated and disaggregated analysis. Renew. Sustain. Energy Rev. 2017, 75, 220–233. [CrossRef] Wang, H.L.; Ou, X.M.; Zhang, X.L. Mode, technology, energy consumption, and resulting CO2 emissions in China’s transport sector up to 2050. Energy Policy 2017, 109, 719–733. [CrossRef] Zhang, N.; Yu, K.; Chen, Z. How does urbanization affect carbon dioxide emissions? A cross-country panel data analysis. Energy Policy 2017, 107, 678–687. [CrossRef] Xi, X.; Qiao, Y.B.; Wu, K.P.; Li, C. The Selection of Metropolises or Towns in Urbanization in the Perspective of Sustainable Development: Based on the Research of the International Ecological Footprint Panel Data. China Popul. Resour. Environ. 2015, 25, 47–56. Bao, J.Q.; Yu, T.Y. Zhejiang should take the lead in meeting the “carbon emissions peak”. Zhejiang Econ. 2015, 12, 26–27. Tapio, P. Towards a theory of decoupling: Degrees of decoupling in the EU and the case of road traffic in Finland between 1970 and 2001. Transp. Policy 2005, 12, 137–151. [CrossRef] Gao, W. China’s Energy Consumption Growth Mechanism and Regional Characteristics; Chemical Industry Press: Beijing, China, 2016. Zhejiang Statistical Yearbook; China Statistics Press: Beijing, China; 1990–2017. (In Chinese). China Statistical Yearbook; China Statistics Press: Beijing, China; 1990–2017. (In Chinese). Yang, Y.L.; Zhao, R.J. Study on the Evaluations System of Industrial Structure Rationalization in China. J. Ind. Technol. Econ. 2017, 1, 62–65. Wang, Y.; Cui, Y.T.; Wang, Z.Q. A comparative study of the industrial structure characteristics of China’s provinces-based on the perspective of industrial space structure. Mod. Manag. Sci. 2016, 3, 6–8. Zhou, Y.X.; Tian, S. Data Revision of Urbanization Level of China’s provinces Based on “The fifth census” Data. Stat. Res. 2017, 36, 74–82. Zellner, A.; Palm, F.C. Time Series Analysis and Simultaneous Equation Econometric Models. J. Econ. 1974, 2, 17–54. [CrossRef] Sims, C. Macroeconomics and Reality. Econometrica 1980, 48, 1–48. [CrossRef] Amisano, G.; Giannini, C. Topics in Structural VAR Econometrics; Springer: New York, NY, USA, 1997. Soccorsi, S. Measuring nonfundamentalness for structural VARs. J. Econ. Dyn. Control 2016, 71, 86–101. [CrossRef] Luetkepohl, H.; Milunovich, G. Testing for identification in SVAR-GARCH models. J. Econ. Dyn. Control 2016, 73, 241–258. [CrossRef] Watson, M.W. Vector Autoregressions and Cointegration. In Handbook of Econometrics; Engle, R.F., McFadden, D., Eds.; North Holland: Amsterdam, The Netherlands, 1994.

Sustainability 2018, 10, 1889

43. 44.

12 of 12

Gao, T.M. Econometric Methods and Modeling-Eviews Applications and Examples, 2nd ed.; Tsinghua University Press: Beijing, China, 2009. Fu, K.; Allen, M.; Archibald, R. Evaluating the Relationship between the Population Trends, Prices, Heat Waves, and the Demands of Energy Consumption in Cities. Sustainability 2015, 7, 15284–15301. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).